Prompt Caching（提示词缓存）

对重复使用的长提示词开启缓存，后续请求费用降低约 90%，延迟降低约 85%。

Python

import anthropic

client = anthropic.Anthropic(api_key="your-api-key", base_url="https://www.cheapertoken.work")

with open("long_document.txt") as f:
    document = f.read()

response = client.messages.create(
    model="claude-opus-4-7",
    max_tokens=1024,
    system=[
        {"type": "text", "text": "你是一个文档分析助手。"},
        {
            "type": "text",
            "text": document,
            "cache_control": {"type": "ephemeral"}
        }
    ],
    messages=[{"role": "user", "content": "请总结这篇文档的主要内容"}]
)

print(response.content[0].text)
print(f"缓存创建 token: {response.usage.cache_creation_input_tokens}")
print(f"缓存命中 token: {response.usage.cache_read_input_tokens}")

JavaScript

import Anthropic from "@anthropic-ai/sdk";
import fs from "fs";

const client = new Anthropic({ apiKey: "your-api-key", baseURL: "https://www.cheapertoken.work" });

const document = fs.readFileSync("long_document.txt", "utf-8");

const response = await client.messages.create({
  model: "claude-opus-4-7",
  max_tokens: 1024,
  system: [
    { type: "text", text: "你是一个文档分析助手。" },
    { type: "text", text: document, cache_control: { type: "ephemeral" } },
  ],
  messages: [{ role: "user", content: "请总结这篇文档的主要内容" }],
});

console.log(response.content[0].text);
console.log(`缓存创建 token: ${response.usage.cache_creation_input_tokens}`);
console.log(`缓存命中 token: ${response.usage.cache_read_input_tokens}`);

curl

curl https://www.cheapertoken.work/v1/messages \
  -H "x-api-key: your-api-key" \
  -H "anthropic-version: 2023-06-01" \
  -H "content-type: application/json" \
  -d '{
    "model": "claude-opus-4-7",
    "max_tokens": 1024,
    "system": [
      {"type": "text", "text": "你是一个文档分析助手。"},
      {"type": "text", "text": "<your long document here>", "cache_control": {"type": "ephemeral"}}
    ],
    "messages": [{"role": "user", "content": "请总结这篇文档的主要内容"}]
  }'

多轮对话中缓存工具定义

Python

import anthropic

client = anthropic.Anthropic(api_key="your-api-key", base_url="https://www.cheapertoken.work")

tools = [
    {
        "name": "search",
        "description": "搜索知识库",
        "input_schema": {
            "type": "object",
            "properties": {"query": {"type": "string"}},
            "required": ["query"]
        },
        "cache_control": {"type": "ephemeral"}
    }
]

response = client.messages.create(
    model="claude-opus-4-7",
    max_tokens=1024,
    tools=tools,
    messages=[{"role": "user", "content": "搜索关于Python的内容"}]
)
print(response.content[0].text)

JavaScript

import Anthropic from "@anthropic-ai/sdk";

const client = new Anthropic({ apiKey: "your-api-key", baseURL: "https://www.cheapertoken.work" });

const tools = [
  {
    name: "search",
    description: "搜索知识库",
    input_schema: {
      type: "object",
      properties: { query: { type: "string" } },
      required: ["query"],
    },
    cache_control: { type: "ephemeral" },
  },
];

const response = await client.messages.create({
  model: "claude-opus-4-7",
  max_tokens: 1024,
  tools,
  messages: [{ role: "user", content: "搜索关于Python的内容" }],
});
console.log(response.content[0].text);

curl

curl https://www.cheapertoken.work/v1/messages \
  -H "x-api-key: your-api-key" \
  -H "anthropic-version: 2023-06-01" \
  -H "content-type: application/json" \
  -d '{
    "model": "claude-opus-4-7",
    "max_tokens": 1024,
    "tools": [{
      "name": "search",
      "description": "搜索知识库",
      "input_schema": {"type": "object", "properties": {"query": {"type": "string"}}, "required": ["query"]},
      "cache_control": {"type": "ephemeral"}
    }],
    "messages": [{"role": "user", "content": "搜索关于Python的内容"}]
  }'

缓存最短有效期为 5 分钟，最长 1 小时。最少需缓存 1024 个 token（Haiku 为 2048）。